Proc. Natl. Acad. Sci. USA Vol. 89, pp. 4338-4342, May 1992 Cell Biology SNCI, a yeast homolog of the synaptic vesicle-associated membrane /synaptobrevin family: Genetic interactions with the RAS and CAP (signal transduction/onogenes/vesicular tafficking/cAMP) JEFFREY E. GERST*, LINDA RODGERSt, MICHAEL RIGGSt, AND MICHAEL WIGLERtI *Department of Cell Biology and Anatomy, Mount Sinai School of Medicine, One Gustave Levy Place, New York, NY 10029-6574; and tCold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, NY 11724 Contributed by Michael Wigler, February 7, 1992

ABSTRACT SNCI, a gene from the yeast Saccharomyces MATERIALS AND METHODS cerevisiae, encodes a homolog of vertebrate synaptic vesicle- associated membrane (VAMPs) or synaptobrevins. Microbial Culture. Yeast strains were grown in rich medium SNCI was isolated by its ability to suppress the loss of CAP (YPD; yeast extract/Bactopeptone/dextrose), synthetic com- (SC), or SC drop-out minimal medium function in S. cerevisiae strains possessing an activated of plete minimal medium lacking an essential amino acid or nucleotide base. Drop-out is a ofthe RAS-responsive S. cerevisiae RAS2. CAP component minimal medium was used to maintain selection of plasmids. adenylyl cyclase complex. The N-terminal domain of CAP is Yeast extract, Bactopeptone, and yeast nitrogen base lacking required for full cellular responsiveness to activated RAS ammonium sulfate and amino acids (YNB) were purchased proteins. The C-terminal domain of CAP is required for from Difco. YPD was prepared according to Sherman et al. normal cellular morphology and responsiveness to nutrient (17). SC minimal complete and drop-out media were prepared extremes. Multicopy plasmids expressing SNCI suppress only as described by Sherman et al. (17) and consisted of0.7% YNB the loss of the C-terminal functions of CAP and only in the supplemented with the appropriate auxotrophic requirements presence of activated RAS2. and 2% glucose. Yeast medium lacking in amino acids and a nitrogen source (YNB-N) was prepared according to Toda et The yeast Saccharomyces cerevisiae contains two RAS al. (4). Standard methods were used to introduce plasmids into genes that encode proteins highly homologous to mammalian the various yeast strains (17). Escherichia coli strains HB101 RAS oncogene products (1, 2). These RAS proteins are and DH5a were used for plasmid transformations and plasmid required to activate S. cerevisiae adenylyl cyclase (3, 4) but DNA preparations. may have other functions as well (5). The functions of RAS Yeast Strains. cap yeast strains SKN32 (Mata leu2 ura3 in higher organisms are not known. When expressed in S. trp) ade8 can) cap::HIS3) and SKN37 (Mata leu2 ura3 trp) cerevisiae, mammalian RAS proteins are capable of both ade8 can) RAS2V1l19 cap::HIS3) have been described (9). activating adenylyl cyclase and suppressing the lethality This cap:.HIS3 allele lacks amino acids 78-451 ofthe coding associated with the loss ofendogenous RAS function (3, 5, 6). region of CAP and is a null allele. cap strain SK013 (Mata Thus, some functions of RAS may have been conserved leu2 ura3 trp) ade8 can) cap::HIS3) has also been described during the course ofevolution. To explore this we have begun (9). cap strain SKN50 (Mata leu2 trp) ade8 can) iral::HIS3 to characterize the S. cerevisiae adenylyl cyclase complex. cap:: URA3) was created by transforming the iral strain IR-1 We previously identified a protein called CAP that copu- (18) with the EcoRI fragment of the CAP disruption plasmid rifles with a RAS-responsive adenylyl cyclase complex (7). pUSMN2 as described (9). cap strains SKN55 and SKN56 The gene for CAP encodes a 526-residue protein that is (Mata ade8 can) bcy)::LEU2 tpk2::HIS3 tpk3::TRP) required for full cellular responsiveness to activated RAS and cap:: URA3) were created by transforming the bcy) tpk2 tpk3 for normal cellular morphology and responsiveness to nutri- strain S13-58A (19) with the EcoRI fragment of pUSMN2 ent extremes (8, 9). Deletion analysis has shown that CAP is CAP disruption plasmid. cap strains SKN58 and SKN59 bifunctional (10). Expression of a domain consisting of the (Mata ade8 trp) can) pdel::LEU2 pde2::URA3 cap::HIS3) N-terminal 168 amino acids is sufficient for full cellular were created by transforming the pdel pde2 strain DJ23-3C responsiveness to activated RAS, while expression of the (20) with the EcoRI fragment of the CAP disruption plasmid C-terminal 160 amino acids is sufficient for normal cellular pHSPN5 as described (9). HIS3+ transformants were isolated responses to nutrient extremes (10). At present, it is unclear and the disruption of CAP was verified by phenotypic and whether RAS or adenylyl cyclase influences CAP function. Southern blot analysis. The snc) strain JG4 (Mata leu2 trp) To understand the function ofCAP, we have isolated genes ade8 his3 can) snc):: URA3) was constructed by transform- that on multicopy plasmids are capable of suppressing loss of ing the haploid SP1 yeast strain (Mata leu2 ura3 trp) ade8 C-terminal function. One such gene, PFY, encodes profilin, his3 can)) (4) with the Sal I/Sac I fragment of the SNCI an actin binding protein (11). Another gene, which we have disruption plasmid pORF3U (see below). Integration of this named SNC) (suppressor of the null allele of CAP), is construct at the SNC) results in an insertion in SNC). described here. It encodes a protein homologous to low The snc) strain JG5 (Mata leu2 trp) ade8 his3 can) snc)A::URA3) was created by transforming the haploid SP1 molecular weight proteins known as VAMPs (synaptic ves- yeast strain with the Sal I/Sac I fragment of the SNC) icle-associated membrane proteins) (12, 13) or synaptobre- of this vins (14-16) that are associated with synaptic vesicles and are disruption plasmid pNCSU (see below). Integration found in a wide variety of organisms.§ Abbreviations: VAMP, synaptic vesicle-associated membrane pro- tein; ORF, open reading frame. The publication costs of this article were defrayed in part by page charge *To whom reprint requests should be addressed. payment. This article must therefore be hereby marked "advertisement" §The sequence reported in this paper has been deposited in the in accordance with 18 U.S.C. §1734 solely to indicate this fact. GenBank data base (accession no. M91157). 4338 Downloaded by guest on September 29, 2021 Cell Biology: Gerst et al. Proc. Natl. Acad. Sci. USA 89 (1992) 4339 construct at the SNCJ locus results in deletion of the SNCI pADH-CAPA15, which expresses 283 amino acids of the N gene. Diploid yeast, formed by mating DC124 (Mata leu2 terminus of CAP (10); and YEpIRA2, which expresses IRA2 ura3 trpl ade8 his4) (19) and SP1, were also transformed with (R.-M. Ballester and M.W., unpublished results). the Sal I/Sac I fragment of pORF3U. Ura' transformants Other plasmids included the following: YEpSNC1, a were sporulated and subjected to tetrad analysis (17). Dis- YEp13M4 plasmid bearing a 3.4-kilobase (kb) Sau3A partial tribution ofthe URA3 marker in haploid strains derived from digestion fragment of genomic SNC1; YEpTSNC1, a pTV3 tetrad analysis was found to be 2:2. In all the cases described plasmid bearing this gene as a Sal I/Sac I fragment; pUC- above, genotypes were verified by Southern blot analysis. SNC, a pUC118 plasmid bearing this fragment; pADH-SNC1 DNA Manipulations. DNA restriction endonucleases, Taq and pADH-cSNC1, which contain a 550-base-pair (bp) frag- polymerase, and T4 DNA ligase were used as recommended ment ofgenomic SNC1 or a 370-bp fragment ofSNCI cDNA, by the suppliers (New England BioLabs and Cetus). Molec- cloned into the Sal I and Sac I sites of pAD4A, respectively; ular cloning, Southern blotting, and colony hybridization pADH-ASNC1, which bears a 2.3-kb BamHI fragment, con- techniques were performed as described by Maniatis et al. taining the ADHI promoter and the SNCJ sequence from (21). DNA sequencing was performed by the dideoxynucle- pADH-SNC1, cloned into the BamHI site of pAV3; and otide chain-termination method (22). The polymerase chain pADH-ACAPA4, which bears a 3.0-kb BamHI fragment, reaction (PCR) (23) and subcloning of PCR products were containing the ADHI promoter and the CAPA4 allele (10) carried out as described (10). Oligonucleotides used to am- from pADH-CAPA4, cloned into the BamHI site of pAV3. plify SNCJ included a forward oligonucleotide bearing a Sal The modified pAV3 plasmids were used in the transformation I site (5'-AACGTATTCGTCGACCATGTCGTC-3') and a of cap strains SKN55-56 and SKN58-59. reverse oligonucleotide bearing a Sac I site (5'-CTA- Two plasmids derived from pUCSNC were used for dis- CATATGGGAGCTCCCTAT-3'). Total RNA was isolated ruptions of SNCI: pORF3U, which has the URA3 selectable from wild-type yeast (SP1) according to Sherman et al. (17). marker cloned into the Sty I site ( 325) of SNC1; and Isolation ofpoly(A)+ RNA was accomplished with a kit from pNCSU, which has URA3 cloned into the Spe I sites, which Stratagene. First-strand cDNA synthesis from yeast flank SNCJ (base pairs -237 and +646, respectively). For poly(A)+ RNA, which was used as a template for PCR, was pNCSU, pUCSNC was digested with Spe I, which removes accomplished with a cDNA synthesis kit from Bethesda the entire SNCJ coding region, and the URA3 gene was Research Laboratories. cloned into this site. Both disruption constructs were verified Genomic DNA from a YPD' revertant of the SK013 cap by restriction analysis. mutant strain (9) was isolated according to Sherman et al. (17). The DNA was partially digested with the Sau3A re- RESULTS striction endonuclease and size-fractionated by gel electro- A Yeast Gene That Suppresses Loss of the C-Terminal phoresis. A library was constructed in the yeast expression Functions of CAP. We have screened a yeast genomic library plasmid YEp13M4 (20) from size-selected DNA cloned into for multicopy suppressors of the cap null of the the BamHI site of the vector. SKN37 strain. This strain is unable to grow on the standard Phenotypic Assays and Selections. cap strain SKN37 was amino acid-rich medium YPD. Expression of the C-terminal transformed with the library and grown for 48 hr on selective domain ofCAP is fully able to complement this defect (10). Of medium before replica plating onto YPD plates to assay for of on YPD in growth on rich medium. Of2000 Leu+ transformants, 30 were 2000 transformants only one was capable growth YPD+, only one ofwhich was found to grow on rich medium a plasmid-dependent manner. This transformant contained in a plasmid-dependent manner. Plasmids were isolated from YEpSNC1, a plasmid that was found to be as potent a yeast according to standard methods (17). Candidate plas- suppressor of rich medium growth defects in SKN37 cells as mids were then retransformed back into the cap SKN37 plasmids expressing either full-length CAP or the C-terminal strain and were examined for their ability to confer growth on domain of CAP (Fig. 1). SKN37 cells display other defects YPD. Assays for growth on rich medium, temperature- associated with loss of the C-terminal function of CAP, sensitive growth at 37°C, and sensitivity to nitrogen starva- including large cell size, temperature-sensitive growth, and tion or heat shock (55°C) were performed as described (10). inability to withstand nitrogen starvation (9). YEpSNC1 sup- Plasmids. Plasmids used in this study included the following pressed the abnormal cell size (Fig. 2) and the other growth vectors: YEp13M4, yeast expression plasmid bearing the defects as well as did plasmids expressing the C-terminal LEU2 selectable marker (20); pAD4A&, a similar plasmid bear- domain of CAP (data not shown). In contrast, YEpSNC1 did ing both the LEU2 selectable marker and the ADHI promoter not suppress loss of the N-terminal functions of CAP, the loss (18); and pUV2, pTV3, and pAV3, YEp-based plasmids of full RAS responsiveness. Wild-type yeast expressing the bearing the URA3, TRPI, or ADE8 selectable markers, re- activated RAS2Va`L9 allele are sensitive to heat shock (24). spectively. Previously described plasmids included the fol- SKN37 cells, which contain the activated RAS allele, are lowing: pADH-CAP, which expresses full-length CAP under resistant to heat shock because of the absence of the N-ter- the control of the ADHI promoter (9); pADH-CAPA&4, which minal domain of CAP (10). SKN37 cells containing the YEp- expresses 237 amino acids of the C terminus of CAP (10); SNC1 plasmid remain resistant to heat shock (data not shown). SC YPD Strain Plasmid 300C 3C0c __-. SPI (CAPWt) YEp 13M4 FIG. 1. Effect of SNCI on the growth of cap strains on rich medium. cap yeast strain SKN37 was transformed with yeast episomal plasmids ex- SKN 37 (Scg RAS2VaI119) YEp13M4 pressing the C-terminal domain of CAP (pADH- CAPA4), the N-terminal domain of CAP (pADH- SKN37 pADH-CAPA4 CAPA15), full-length CAP (pADH-CAP), or SNC1 SKN 37 pADH-CAPA15 (YEpSNC). Both CAP' (SP1) and cap yeast were transformed with an empty vector (YEp13M4) as SKN 37 pADH- CAP control. Patches of transformed yeast were grown SKNS7 YEpSNC for 3 days on synthetic minimal medium (SC) before replica plating to rich growth medium (YPD). Downloaded by guest on September 29, 2021 4340 Cell Biology: Gerst et al. Proc. Natl. Acad. Sci. USA 89 (1992) 0

(~~~~~. ..

2i . FIG. 2. Effect of SNCI on the n ffi sr. :. 0 n morphology of cap yeast. cap yeast ) SO. ':'.. .:' ....: .' ;:.. is strain SKN37 (RAS2VAl9 cap:.HIS3) ...... ': .... : .l was transformed with eithera plasmid *. :..: .. . ( expressing SNC1 (pADH-SNC1) or ,: i: a:. ..sX :,,.2J 4sW 2 YEp13M4 as a control. CAPWt ..,...., :i. ,,;F,.,..%"S.,,9..,I'",, R I,E RAS2Va'l9 yeast were also trans- formed with YEp13M4 as a control. Transformed yeast strains were Val 19 wt vcl 19 V RAS? CAP RAS? c::p>HiS RAS2 op:: H17-,Z grown in liquid synthetic medium for 2 days before visual inspection. onritr!! Controi pAD A - Sheik (x 1050.) Sequence of SNC1. The YEpSNC1 plasmid contained an Homology Between SNC1 and Mammalian VAMPs. The insert of3.4 kb. Deletion analysis ofthe plasmid localized the yeast SNC1 protein is highly homologous to both the type 1 full suppressor activity of YEpSNC1 to a 1-kb fragment. and type 2 mammalian VAMP/synaptobrevins (Fig. 4). The DNA sequencing of this fragment revealed two small open amino acid sequence of SNC1 shows =40%6 identity and reading frames (ORFs) (Fig. 3). A search for similar DNA =50% overall homology to either type of protein. The sequences revealed a strong homology between these ORFs, VAMP/synaptobrevin protein has been structurally divided when taken together as a single ORF, and members ofa class into three distinct regions based on analysis ofthe secondary of small, highly conserved proteins [VAMPs (12, 13) or structure and tryptic digestion patterns (12-14, 16). The synaptobrevins (14-16)]. These proteins have been identified N-terminal 80-90 amino acids constitute a highly charged in a wide variety of organisms (from fish to humans). Their hydrophilic region that extends into the cytosol (13). The first function is unknown, but they are suspected to play a role in 20 residues of the N terminus of both the type 1 and type 2 vesicle fusion and neurotransmitter release. Two classes of proteins are highly diverged (12, 14). This is the region with these proteins have been identified-the type 1 and type 2 the least homology to yeast SNC1. C-terminal to the cytosolic VAMP/synaptobrevins, which differ slightly in their N-ter- domain is a stretch of 20-25 hydrophobic residues that minal domains. constitute a transmembrane domain. The last four residues of Alignment between the translated sequence of the ORFs the C terminus of the protein are thought to constitute an and those of the VAMP/synaptobrevin genes suggested that, intravesicular domain (12-16). The region ofstrong homology together, these ORFs constitute the entire coding region of a between SNC1 and the VAMP/synaptobrevins lies down- yeast homolog. Several lines of evidence confirm this. First, stream of residue 20 and extends throughout the hydrophilic SNCI contains sequences that match the consensus sequences region (residues 21-91) and includes the stretch of -20 for splice donor and acceptor sites in S. cerevisiae: the 5' splice hydrophobic amino acids. Computer analysis by the method motif (GTANGT) and the 3' splice motif (YAG, where Y is of Klein et al. (27) predicts the SNC1 protein to be an integral pyrimidine) (25) are found at the candidate intron-exon junc- membrane protein. The membrane-spanning region consists tion at base pairs 101-106 and 213-215, respectively. In of residues 95-111. addition, the TACTAAC box, a consensus motifpresent in S. Disruption of the SNC1 Locus. The chromosomal SNCJ cerevisiae introns (26), is found at base pairs 146-152. Second, locus was disrupted by insertion into, ordeletion of, the coding oligonucleotide-directed amplification of SNC1 from yeast region of SNCJ by transformation (see Materials and Meth- cDNA by PCR generated a product =100 bp smaller than that ods). Microscopic and phenotypic analysis ofconfirmed sncl generated when genomic DNA was used as a template (data strains indicated that the disruption of SNC1 has no apparent not shown). DNA sequencing of the PCR product generated phenotype in yeast (data not shown). sncl strains were found from yeast cDNA revealed that base pairs 101-215 were to grow and mate in a normal fashion, were morphologically absent. Finally, expression ofthe PCR-amplified cDNA clone normal, and did not show that result from loss of was able to suppress the CAP-deficient phenotypes of SKN37 the C-terminal functions ofCAP. Haploid sncl strains derived cells (data not shown). These studies demonstrate not only the from tetrad analysis also had no phenotype. Thus, SNCI does existence of an intron but also that we have identified the not, by itself, provide an essential function. correct ORF with suppressor activity. Genetic Interactions Between SNC1 and RAS2. SNCI was The ATG codon starting at nucleotide 1 is preceded by isolated as a suppressor of cap phenotypes in the yeast multiple in-frame stop codons. Given this, we conclude that SKN37 strain, which contains an activated RAS2Va`l9 allele. SNCJ encodes a protein of 117 amino acids and has a We therefore determined whether SNC1 was also capable of predicted molecular mass of =13 kDa. A 115-bp intron suppressing cap defects in the cap RAS2' strains SKN32 and sequence connects the region encoding the first 33 amino SKN34. In contrast to cells transformed with plasmids ex- acids to that encoding the last 84 amino acids. pressing either full-length CAP or the C-terminal domain of I ATG TCG TCA TOT ACT CCC TTT GAC CCT TAT GCT CTA TCC GAG CAC GAT GAA GAA CGA M S S S T P F D P Y A L S EH D EE FIG. 3. Sequence of the SNCI CCC CAG AAT GTA CAG TCT AAG TCA AGG ACT GAA CTA CAA OCT GTAAGTACAGAAAGC PQ N V QSK S R T AC0A EL A gene and its encoded product. Coor- 118 CACAGAGTACCATCTAGGAAATTAACATTATaCTA&CTTJ dinates for nucleotides and codons 13 ACGTTCTTCGTGTTTATTTTTAG GAA ATT GAT GAT ACC GTG GGA ATA ATG AGA GAT AAC ATA are indicated on the left. The single- E I D D T V G I M R D N I letter amino acid code for the SNC1 255 AAT AAA GTA GCA GAA AGA GAA TTA ACG ATT GA GDAT AAA GCC GAT AAC 48N K V A E R GSTG E ARAR L TSC K A D N protein is given below each 3-bp 3jj CTA GAG GTC TCA GCC CAA GGC TTT AAG AGG GGT GCC AAT AGG GTC AGA AAA GCC ATG codon of the ORFs. The consensus IL A V A F K R GA N R V R K A M sequences for intron splicing in S. TGO TAC AAG GAT CTA AAA ATG AAG ATG T8T CTG GOT TTA GTA ATC ATC ATA TTG CTT cerevisiae are indicated in boldface 31g L A L V I I I L L 426 GTT GTA ATC ATC GTC CCC ATT GCT GTT CAC TTT CGA TAG type. The stop codon is indicated with 10 V V I I V P I A V H AWT an asterisk. Downloaded by guest on September 29, 2021 Cell Biology: Gerst et al. Proc. Natl. Acad. Sci. USA 89 (1992) 4341

RATVAMP1 1 MSAPA--QPPAEGTE--GAAPGGGPPGPPPNTTSNRRLQQTQAQVEEVVDIIRVNVD 11 11 ::: 1:1 1: Yeast SNC1 1 MSSSTPFDPYALSEHDEER------PQNVQSKSRTAELQAEIDDTVGIMRDNIN 11: I I 1 1: 11 11:1:I1 1: RATVAMP2 1 MSATAATVPPA------APAGEGGPPAPPPNLTSNRRLQQTQAQVDEVVDIMRVNVD FIG. 4. Comparison between SNC1 and the VAMP/synaptobrevin proteins. The amino acid sequence of SNC1 was aligned with the two types of rat VAMP/ 53 KVLERDQKLSELDDRADALQAGASVFESSAAKLKRKYWWKNCKMMIMLGAICAIIVVVIVI----YIFT. synaptobrevin proteins. Vertical bars 11 11 :1: ::I:11 I :::: 1:1 11: 1 : 1::111:: designate identities between pairs; dots 48 KVAERGERLTSIEDKADNLAVSAQGFKRGANRVRKAMWYKDLKMKMCLALVIIILLVVIIVPIAVHFSR. represent conservative amino acid sub- 11 11 :I: ::I:11I I :::: 1:1 111: :: 1:1::111 I I stitutions. Amino acid coordinates are on 53 KVLERDQKLSELDDRADALQAGASQFETSAAKLKRKYWWKNLKMMIILGVICAIILIIIIV----YTST. the left. CAP, SKN32 or SKN34 cells transformed with YEpSNC1 functions of CAP-namely, growth at 37°C (Fig. 5B) and remained incapable ofgrowth on rich medium or on synthetic resistance to nitrogen starvation. minimal medium at 37TC and retained their abnormal mor- To verify further that the functional activity of SNCI phology and sensitivity to nitrogen starvation. These results results from activation of RAS, we coexpressed the yeast suggest that activation of the RAS pathway is required for IRA2 gene from a multicopy plasmid containing the LEU2 SNCI suppressor activity. selectable marker YEpIRA2 and SNC1 from a multicopy To rule out the possibility that undetected genetic differ- plasmid containing the TRPI selectable marker YEpTSNC1 ences between the cap RAS21 strains (SKN32 and SKN34) in SKN50 cells. The IRA2 gene encodes a second GAP-like and the cap RAS2Va`l9 strain (SKN37) might be affecting the protein and is able to functionally complement the loss of suppressor activity of SNCJ, we created another cap null IRA) (30). SKN50 cells transformed with both SNC1 and strain, SKN50 (see Materials andMethods). This strain bears IRA2 plasmids were found to grow slowly on YPD and were the RAS2Wt allele but also has a disrupted IRA] locus. The unable to grow altogether at 37°C under conditions selective yeast IRA] gene is a functional homolog of the mammalian for the plasmids (Fig. 5B). In contrast, SKN50 cells trans- GTPase-activating protein (GAP) (18, 28). An iral CAP' formed with the SNCI plasmid and a control plasmid bearing strain shows all the phenotypes associated with persistent the LEU2 marker YEp13M4 were capable of strong growth activation of the RAS pathway (3, 24, 28), including heat either on YPD or on synthetic medium at 37°C. Thus, the shock sensitivity, presumably due to perpetuation of RAS in overexpression ofIRA2 appears to inhibit the ability ofSNCJ its activated GTP-bound state (29). SKN50 cells, however, to suppress the loss of CAP functions in SKN50 cells. This are resistant to heat shock because of loss of the N-terminal conclusion was further confirmed by plasmid segregation functions of CAP and also show the phenotypes associated analysis (data not shown). with loss of the C-terminal functions of CAP (data not Genetic Interactions Between SNCI and Genes Encoding shown). The typical iral phenotypes can be restored to Components of the cAMP Pathway. Evidently expression of SKN50 cells upon expression offull-length CAP. In contrast SNCI suppresses the phenotype resulting from loss of C-ter- to the results obtained with the SKN32 and SKN34 strains, minal CAP function only in cells that have activated RAS. but like our results with the SKN37 strain, YEpSNC1 was One possible explanation for this phenomenon is that acti- able to confer robust growth on rich medium in SKN50 cells vation ofRAS stimulates the cAMP effector pathway even in (Fig. In addition, the was able the absence of CAP. We therefore sought to determine SA). expression of SNC1 to whether the suppressor activity of SNCI was dependent on correct other defects associated with loss of the C-terminal stimulation of the cAMP effector pathway. We created four A additional cap mutant strains: two, SKN58 and -59, lacked r-; r. Plosmid SC YPD the cAMP phosphodiesterases encoded by PDEI and PDE2 (20, 24); and two, SKN55 and -56, lacked the regulatory SKN32 (cqp RAS2w) YEpSNC subunit of the cAMP-dependent protein kinases encoded by SKN37 (qcp RAS2Vl 19) YEpSNC the BCYI gene (31) and two of the three catalytic subunit SKN50 (cqq RAS2w irl ) YEpSNC eul isoforms encoded by the TPK2 and TPK3 genes (32). Wild- type yeast lacking the PDE genes are unable to hydrolyze SC YPD SC cAMP and are sensitive to heat shock (20). Yeast lacking 300c 300C 37°C B Stra i Plasmids BCYJ, but expressing at least one of the TPK genes, are also SKN50 YEpTSNC YEpl3M4 heat shock sensitive (31). As expected, disruption of CAP in SKN50 YEpTSNC YEp 13M4 these yeast strains had no effect on their sensitivity to heat ml= shock, because the N-terminal function of CAP is not re- SKN 50 YEpTSNC YEplRA2 quired in these strains. Like other cap disruption strains, SKN5 vEPTSNC YEpIRA2 SKN55, -56, -58, and -59 were unable to grow on rich medium and displayed the other defects associated with loss of the FIG. 5. Effect of SNCI in cap strains having a disruption of the C-terminal functions of CAP. Overexpression of SNCJ was IRAI locus. (A) cap strains SKN32 (RAS2'), SKN37 (RAS2VaIS9), unable to confer growth on rich medium in these four cap and SKN50 (RAS21 iral) were transformed with a plasmid (YEp- mutant strains. On the basis of these experiments, it appears SNC) expressing SNC1. These strains were also transformed with an that the functional activity of SNC1 in cap mutant strains is empty vector (YEp13M4) as control (data not shown). After growth dependent on the activated state of RAS but not on the level for 3 days on synthetic medium (SC), patches of transformed yeast of cAMP or cAMP-dependent protein kinase activity. were replica plated onto YPD plates to test for growth on rich medium (YPD). (B) cap strain SKN50 (RAS21 iral) transformed DISCUSSION with a plasmid (YEpTSNC) expressing SNC1 was transformed with a second plasmid (YEpIRA2) expressing either IRA2 or a control The SNCI gene encodes a protein of 117 amino acids that is vector (YEp13M4). Double transformants were grown for 3 days on highly homologous to members ofthe VAMP/synaptobrevin synthetic medium (SC) before replica plating onto YPD plates to test family (12-16), which, because of their subcellular localiza- for growth on rich medium (YPD; 30°C) or to prewarmed synthetic tion, are presumed to be involved in targeting and fusion of medium to test for temperature-sensitive growth at 37°C (SC; 37°C). synaptic vesicles with the presynaptic membrane (12-16). Downloaded by guest on September 29, 2021 4342 Cell Biology: Gerst et al. Proc. Natl. Acad. Sci. USA 89 (1992) These proteins have a distinct structure, composed of a 5. Wigler, M., Field, J., Powers, S., Broek, D., Toda, T., Cam- C-terminal transmembrane domain and a small cytoplasmic eron, S., Nikawa, J., Michaeli, T., Colicelli, J. & Ferguson, K. N-terminal domain. Other proteins of similar overall struc- (1988) Cold Spring Harbor Symp. Quant. Biol. 53, 649-655. ture but with larger cytoplasmic domains are known to 6. Kataoka, T., Powers, S., Cameron, S., Fasano, O., Goldfarb, function in the secretory pathway of S. cerevisiae. SNC1 is M., Broach, J. & Wigler, M. (1985) Cell 40, 19-26. very weakly homologous to these other yeast proteins [e.g., 7. Field, J., Nikawa, J., Broek, D., MacDonald, B., Rodgers, L., Wilson, I., Lerner, R. & Wigler, M. (1988) Mol. Cell. Biol. 8, SLY12/BET1 (33, 34), SLY2 (33), and BOS1 (35)]. 2159-2165. A phenotypic assay for the VAMP/synaptobrevins in yeast 8. Fedor-Chaiken, M., Deschenes, R. & Broach, J. (1990) Cell 61, would aid in evaluation of their function. Unfortunately, 329-340. despite the high degree ofconservation between SNC1 and the 9. Field, J., Vojtek, A., Ballester, R., Bolger, G., Colicelli, J., VAMP/synaptobrevins, expression ofeitherthe type 1 ortype Ferguson, F., Gerst, J., Kataoka, T., Michaeli, T., Powers, S., 2 rat VAMP genes does not complement defects seen in cap Riggs, R., Rodgers, L., Wieland, I., Wheland, B. & Wigler, M. cells (data not shown). Suppression of the loss of SNCI (1990) Cell 61, 319-327. function could in principle provide an assay for VAMP/ 10. Gerst, J., Ferguson, K., Vojtek, A., Wigler, W. & Field, J. synaptobrevin function. Unfortunately, disruption of the (1991) Mol. Cell. Biol. 11, 1248-1257. SNCI locus does not cause a discernible phenotype. Recently, 11. Vojtek, A., Haarer, B., Field, J., Gerst, J., Pollard, T., Brown, we have found clear evidence for a second VAMP/ S. & Wigler, M. (1991) Cell 66, 497-505. synaptobrevin homolog in yeast. Disruption of both homologs 12. Elferink, L., Trimble, W. & Scheller, R. (1989) J. Biol. Chem. may result in a phenotype that would allow 264, 11061-11064. analysis of SNC1 13. Trimble, W., Cowan, D. & Scheller, R. (1988) Proc. Natl. function and also provide a way to use yeast as a model Acad. Sci. USA 85, 4538-4542. organism for study of VAMP/synaptobrevin functions. 14. Archer, B., Ozcelik, T., Jahn, R., Francke, U. & Sudhof, T. SNCI is one of two yeast genes that we have described that (1990) J. Biol. Chem. 265, 17267-17273. were isolated by their ability to complement the cellular 15. Baumert, M., Maycox, P. R. Navone, F., De Camilli, P. & defects associated with loss of the C-terminal functions of Jahn, R. (1989) EMBO J. 8, 379-384. CAP. The first of these, PFY (11), encodes profilin, an 16. Sudhof, T., Baumert, M., Perin, M. & Jahn, R. (1989) Neuron actin/phospholipid binding protein that is presumed to be 2, 1475-1481. involved in cytoskeletal organization (36) but that might also 17. Rose, M. D., Winston, F. & Hieter, P., eds. (1990) Methods in be involved in inositol phospholipid metabolism (37-39). We Yeast Genetics: A Laboratory Course Manual (Cold Spring do not understand how profilin suppresses the loss of CAP Harbor Lab., Cold Spring Harbor, NY), pp. 117-167. function, nor do we understand the mechanism for SNC1 18. Ballester, R., Michaeli, T. Ferguson, K., Xu, H.-P., McCor- suppression. We have postulated that mick, F. & Wigler, M. (1989) Cell S9, 681-686. profilin alters phospho- 19. Cameron, S., Levin, L., Zoller, M. & Wigler, M. (1988) Cell 53, lipid metabolism (11), and it is conceivable that SNC1 may 555-566. directly or indirectly alter phospholipid content or distribution 20. Nikawa, J., Sass, P. & Wigler, M. (1987) Mol. Cell. Biol. 7, within cellular compartments. We are unlikely to resolve this 3629-3636. problem without understanding the normal function of SNC1. 21. Maniatis, T., Fritsch, E. & Sambrook, J. (1982) Molecular We and others (8,9) have shown that yeast with a disrupted Cloning:A Laboratory Manual (Cold Spring Harbor Lab., Cold CAP locus is at least partly uncoupled from activated RAS Spring Harbor, NY). proteins: cells without CAP function do not display the phe- 22. Sanger, F., Nicklen, A. & Coulsen, A. (1977) Proc. Natl. Acad. notypes normally associated with the expression of activated Sci. USA 74, 5463-5467. RAS. Our experiments with SNC1 suggest that cap strains are 23. Saiki, R., Gelfand, D., Stoffe, S., Scharf, S., Higushi, R., Horn, in fact at least partly responsive to activated RAS, although G., Mullis, K. & Ehrlich, H. (1988) Science 239, 487-491. 24. Sass, P., Field, J., Nikawa, J., Toda, T. & Wigler, M. (1986) apparently in a manner independent of the RAS-adenylyl Proc. Natl. Acad. Sci. USA 83, 9303-9307. cyclase interaction. SNC1 is capable of suppressing defects 25. Leer, R., Van Raamsdonk-Duin, M., Hagendorn, M., Mager, associated with loss of the C-terminal function of CAP only in M. & Planta, R. (1984) Nucleic Acids Res. 12, 6685-6690. strains with constitutively activated RAS2 protein. SNCI is 26. Langford, C. & Gallwitz, D. (1983) Cell 33, 519-527. unable to confer suppressor activity in cap strains bearing the 27. Klein, P., Kanehisa, M. & DeLisi, C. (1985) Biochim. Biophys. RAS21 allele, even when other negative regulators of the Acta 815, 468-476. cAMP effector pathway are absent. Thus, the suppressor 28. Tanaka, K., Matsumoto, K. & Toh-e, A. (1989) Mol. Cell. Biol. activity of SNC1 appears to be independent of the presently 9, 757-768. known downstream effectors of RAS in yeast. This result 29. Trahey, M. & McCormick, F. (1987) Science 238, 542-545. 30. Tanaka, K., Nakafuku, M., Tamanoi, D., Kaziro, Y., Matsu- confirms that RAS proteins in the yeast S. cerevisiae have moto, K. & Toh-e, A. (1990) Mol. Cell. Biol. 10, 4303-4313. functions in addition to regulation of adenylyl cyclase (5). 31. Toda, T., Cameron, S., Sass, P., Zoller, M. & Wigler, M. (1987) Moreover, these functions do not depend on CAP. Cell 50, 277-287. We are grateful to Dr. Richard Scheller for providing rat VAMP] 32. Toda, T., Cameron, S., Sass, P., Zoller, M., Scott, J. Mc- and VAMP2 cDNAs and to Peter Novick for a critical reading of the Mullen, B., Murwitz, M., Krebs, E. & Wigler, M. (1987) Mol. manuscript. We wish to thank Patricia Bird for her secretarial Cell. Biol. 7, 1371-1377. assistance. This work was supported by grants from the National 33. Dascher, C., Ossig, R., Gallwitz, D. & Schmitt, H. (1991) Mol. Cancer Institute and the American Cancer Society. M.W. is an Cell. Biol. 11, 872-885. American Cancer Society Research Professor. J.E.G. was supported 34. Newman, A., Shim, J. & Ferro-Novick, S. (1990) Mol. Cell. by a postdoctoral fellowship from the Andrew Seligson Foundation. Biol. 10, 3405-3414. 35. Shim, J., Newman, A. & Ferro-Novick, S. (1991) J. Cell Biol. 1. DeFeo-Jones, D., Scolnick, E., Koller, R. & Dhar, R. (1983) 113, 55-64. Nature (London) 306, 707-709. 36. Pollard, T. & Cooper, J. (1986) Annu. Rev. Biochem. 55, 2. Powers, S., Kataoka, T., Fasano, O., Goldfarb, M., Strathern, 987-1035. J., Broach, J. & Wigler, M. (1984) Cell 36, 607-612. 37. Lassing, I. & Lindberg, U. (1985) Nature (London) 314, 472- 3. Broek, D., Samily, N., Fasano, O., Fujiyama, A., Tamanoi, F., 474. Northup, J. & Wigler, M. (1985) Cell 41, 763-769. 38. Lassing, I. & Lindberg, U. (1988) J. Cell. Biochem. 37, 255- 4. Toda, T., Uno, I., Ishikawa, R., Powers, A., Kataoka, R., 267. Broek, S., Cameron, A., Broach, J., Matsumoto, K. & Wigler, 39. Goldschmidt-Clermont, P., Kim, J., Machesky, L., Rhee, S. & M. (1985) Cell 40, 27-36. Pollard, T. (1991) Science 251, 1231-1233. Downloaded by guest on September 29, 2021